The endeavor to treat HER2-amplified cancers through the inactivation of HER2 has proven more difficult than anticipated. The clinical activities of HER2 kinase inhibitors are weak, and better results with HER2 antibodies occur without the disruption of signaling, likely through immunologic mechanisms. The endeavor to inactivate the oncogenic function of HER2 awaits deeper mechanistic insights. Our work has highlighted the critical role of its partner HER3. Although kinase-inactive, HER3 is a functionally active and requisite partner of HER2 in this disease, and its dynamic regulation by a complex downstream network topology can upregulate HER2-HER3 signaling output 100-fold, undermining the efficacies of all forms of HER2 or HER3-targeted therapies. Effective suppression of the functionally relevant HER2-HER3 tumor driver requires deeper insights into the mechanism of signal generation in the pathologic state of HER2 overexpression. We have conducted a structure-function biochemical analysis of HER2-HER3 transactivation, specifically in the state of HER2 overexpression, revealing deeper insights into the mechanism of signal generation in this disease state. We find that the constitutive phosphorylation of HER3 occurs through interactions that do not require extracellular domain (ECD) activation or dimerization, or even proximity, defying the well described mechanisms for normal physiological signaling in this receptor family. However, HER2-HER3 activation does require kinase domain (KD) interactions, which in lieu of ECD-driven proximation events, are stoichiometrically driven by massive HER2 expression seen in these tumors. The evidence reveals functions in the HER3 KD including a function in its c-lobe as an allosteric activator of HER2 KD, a function in its n-lobe surface pocket, a function in the HER2 KD allosteric receiver site, and an unexpected requirement for the HER2 c-tail in promoting HER3 signaling. These data suggest that the plethoras of ECD-targeting approaches being pursued in the biotech sector are unlikely to yield highly effective therapies for HER2-amplified cancers. Rather it redirects this endeavor towards a focus on targeting kinase domain functions. In the next few years of this project we propose to establish the relevance of these findings to the in vivo tumorigenic growth of HER2- amplified breast cancers. In the first aim we will determine whether the described ligand and ECD-independent constitutive HER2-HER3 signal is the actual driver of tumorigenic growth in vivo. The impact of proximity restricting therapeutic modalities will also be tested. In the second aim we will test several newly discovered strategies targeting the KDs of HER2 and HER3 in the treatment of HER2-amplified cancers in vivo. In the third aim we will explore the more vaguely defined role of the HER2 and HER3 c-tails as either simple signaling substrates or functionally more involved in oligomerization and signal generation. These studies will define the rate-limiting steps for signal generation and tumorigenic growth in cancer cells driven by HER2 overexpression, laying the foundation for the development of highly effective therapies for HER2-amplified cancers.